H N P D I S C U S S I O N P A P E R

About this series...

This series is produced by the Health, Nutrition, and Population Family(HNP) of the World Bank’s Human Development Network. The papersin this series aim to provide a vehicle for publishing preliminary andunpolished results on HNP topics to encourage discussion and debate.The findings, interpretations, and conclusions expressed in this paperare entirely those of the author(s) and should not be attributed in anymanner to the World Bank, to its affiliated organizations or to membersof its Board of Executive Directors or the countries they represent.Citation and the use of material presented in this series should takeinto account this provisional character. For free copies of papers inthis series please contact the individual authors whose name appearson the paper.

Enquiries about the series and submissions should be made directly tothe Managing Editor Rama Lakshminarayanan (rlakshminarayana@worldbank.org) or HNP Advisory Service (healthpop@worldbank.org,tel 202 473-2256, fax 202 522-3234). For more information, see alsowww.worldbank.org/hnppublications.

THE WORLD BANK

1818 H Street, NWWashington, DC USA 20433Telephone: 202 473 1000Facsimile: 202 477 6391Internet: www.worldbank.orgE-mail: feedback@worldbank.org

MDG Oriented Sector and Poverty ReductionStrategies

Lessons from Experience in Health

Mick Foster

October 2005

MDG ORIENTED SECTOR AND POVERTY REDUCTION STRATEGIES

Lessons from Experience in Health

Mick Foster

October 2005

ii

Health, Nutrition and Population (HNP) Discussion Paper This series is produced by the Health, Nutrition, and Population Family (HNP) of the World Bank's Human Development Network. The papers in this series aim to provide a vehicle for publishing preliminary and unpolished results on HNP topics to encourage discussion and debate. The findings, interpretations, and conclusions expressed in this paper are entirely those of the author(s) and should not be attributed in any manner to the World Bank, to its affiliated organizations or to members of its Board of Executive Directors or the countries they represent. Citation and the use of material presented in this series should take into account this provisional character. For free copies of papers in this series please contact the individual author(s) whose name appears on the paper. Enquiries about the series and submissions should be made directly to the Managing Editor, Rama Lakshminarayanan (rlakshminarayana@worldbank.org). Submissions should have been previously reviewed and cleared by the sponsoring department, which will bear the cost of publication. No additional reviews will be undertaken after submission. The sponsoring department and author(s) bear full responsibility for the quality of the technical contents and presentation of material in the series. Since the material will be published as presented, authors should submit an electronic copy in a predefined format (available at www.worldbank.org/hnppublications on the Guide for Authors page). Drafts that do not meet minimum presentational standards may be returned to authors for more work before being accepted. For information regarding this and other World Bank publications, please contact the HNP Advisory Services at healthpop@worldbank.org (email), 202-473-2256 (telephone), or 202-522-3234 (fax). © 2005 The International Bank for Reconstruction and Development / The World Bank 1818 H Street, NW Washington, DC 20433 All rights reserved.

iii

Health, Nutrition and Population (HNP) Discussion Paper

MDG ORIENTED SECTOR AND POVERTY REDUCTION STRATEGIES

Lessons from Experience in Health

Mick Fostera

a Mick Foster Economics, Ltd., Essex, United Kingdom

Paper prepared for High-Level Forum on the Health, Millennium Development Goals

Paris, France - November 14-15, 2005 Abstract: This review is a contribution to the debate on how to achieve and sustain the health-related Millennium Development Goals. It identifies ways in which government planning and budget processes and the management and delivery of external support can be more effective in supporting the goals, based on a review of how the health MDGs are being taken forward in a sample of low-income countries. Keywords: Poverty, Health, Aid, Budgets, Planning Disclaimer: The findings, interpretations and conclusions expressed in the paper are entirely those of the authors, and do not represent the views of the World Bank, its Executive Directors, or the countries they represent. Correspondence Details: Mick Foster, Mick Foster Economics Ltd., 9 Admirals Walk, Chelmsford, Essex, CM1 2XS England, Telephone 44 1245 355031 (office) and 44 07968 190907 (mobile), Email: mickfoster@btconnect.com Website: www.mickfoster.com

iv

v

Table of Contents

ACKNOWLEDGEMENTS ......................................................................................... VII

ABBREVIATIONS......................................................................................................VIII

MDGS AND NATIONAL TARGETS ............................................................................ 2

SUMMARY ........................................................................................................................ 3

SHOULD STRATEGIES BE “NEEDS BASED,” “RESOURCE CONSTRAINED”—OR BOTH? .................................................................................... 3

SUMMARY ........................................................................................................................ 3

FRAMING HEALTH STRATEGIES TO ACHIEVE THE TARGETS .................... 4

SUMMARY ........................................................................................................................ 8

ESTIMATING THE COSTS OF ACHIEVING THE TARGETS .............................. 8

SUMMARY ........................................................................................................................ 9

ARE MACROECONOMIC FRAMEWORKS TOO RESTRICTIVE? ................... 10

SUMMARY ...................................................................................................................... 13

COORDINATING HEALTH PLANS WITH THE MTEF & THE BUDGET........ 15

SUMMARY ...................................................................................................................... 16

ABSORPTIVE CAPACITY .......................................................................................... 16

SUMMARY ...................................................................................................................... 17

REFORMING DEVELOPMENT ASSISTANCE....................................................... 18

SUMMARY ...................................................................................................................... 20

APPENDIX A: GOALS, TARGETS, AND INDICATORS FOR MDGS 1 THROUGH 7................................................................................................................... 21

APPENDIX B: MDGS AND NATIONAL GOALS IN CASE-STUDY COUNTRIES........................................................................................................................................... 22

APPENDIX C: OFFICIAL DEVELOPMENT AID.................................................... 27

BIBLIOGRAPHY........................................................................................................... 29

vi

List of Tables

Table 1: Trends in Public Expenditures on Health ........................................................5

Table 2: IMF Macroeconomic Frameworks in Case-Study Countries ........................14

Table C1: What percentage of Official Development Aid is recorded in the balance of payments? ....................................................................................................................27

Table C2: What percentage of official flows supports government expenditure? ......28

Table C3: Budget Support as Percentage of Official Finance of Government Expenditure......................................................................................................................................28

List of Boxes

Box 1: Is there a need to delegate more authority to those responsible for achieving results? ...........................................................................................................................7

Box 2: Where does all the aid go? ...............................................................................19

vii

ACKNOWLEDGEMENTS

This HNP Discussion Paper is a summary of “MDG-Oriented Sector and Poverty Reduction Strategies: Lessons from Experience in Health: Main Report” by Mick Foster (2004). The work was carried out with the guidance of a steering group composed of Pablo Gottret, Katja Janovsky, Andrew Cassels, Rebecca Dodd, and Emma Jeffreys. Pablo Gottret contributed significantly to the design and commented and guided throughout. Larissa Pelham provided support with literature search. Qiu Fang provided excellent and extensive support with data analysis. The author is grateful to the World Bank for publishing this report as an HNP Discussion Paper. We are grateful to the many staff members of the World Bank, the International Monetary Fund (IMF), the World Health Organization (WHO), and the Millennium Project who provided us with insights, reports, and data. We are especially indebted to Makiko Harrison for access to the World Bank database on the treatment of the MDGs, to Rebecca Dodd for similar material from the WHO, to the IMF Independent Evaluations Office and the Bank’s Operations Evaluation Department for access to country studies from their evaluation of PRSPs. This report is based on country case studies by Mary Betley (Cambodia, Tajikistan), Paul Smithson (Ethiopia, Ghana), Meg Braddock (Nicaragua), Olivier Appaix (Burkina Faso), Eka Ikpe (Uganda, Rwanda), Warren Benfield (Guyana), Mohit Chaturvedi (Nepal, Vietnam), Astrid Cox (Benin), and Mick Foster (Albania, Tanzania). Mick Foster is responsible for errors of fact and judgment in this report. The original draft of this document was presented at the High-Level Forum (HLF) on Health MDGs in Abuja in December 2004. The HLF secretariat co-sponsored the development of the document.

viii

ABBREVIATIONS

ARI Acute respiratory infection CMR Child mortality rate CPIA Country Policy and Institutions Assessment GFATM Global Fund to Fight AIDS, Tuberculosis and Malaria HAART Highly active antiretroviral therapy IDA International Development Association IEO Independent Evaluations Office, International Monetary Fund IMF International Monetary Fund IMR Infant mortality rate ITN Insecticide-treated bednet JSA Joint Staff Assessment (of a PRSP) HIPC Highly Indebted Poor Countries HLF High-Level Forum IMR Infant mortality rate MBB Marginal budgeting for bottlenecks MCH Maternal and child health MDG Millennium Development Goals MOH Ministry of Health MMR Maternal mortality rate MTEF Medium-Term Expenditure Framework MP Millennium Project NDP National Development Plan ODA Official Development Assistance OECD Organization for Economic Cooperation and Development OED Operations Evaluation Department, World Bank PCR Primary completion rate PRGF Poverty Reduction and Growth Facility PRSC Poverty Reduction Support Credit PRSP Poverty Reduction Strategy Paper UPE Universal primary education

1

Eight internationally agreed Millennium Development Goals (MDGs) to be achieved by 2015 were set forth in the Millennium Declaration adopted by the United Nations General Assembly in September 2000. Four of them are related to health:

• Reducing under-five mortality by two thirds • Reducing maternal mortality by three quarters • Reversing the spread of HIV/AIDS, malaria, TB and other communicable

diseases • Reducing the proportion of people who suffer from hunger by half

The goals were not new. They had been discussed and endorsed at many international forums throughout the 1990s. Yet progress on health bogged down in the 1990s because of the AIDS epidemic, the resurgence of diseases such as malaria and TB, and generally dysfunctional health services almost everywhere (UNDP n.d.). In Sub-Saharan Africa, these health-related factors were compounded by slow economic growth and falling per capita incomes, and by problems of war and insecurity in some major countries. Since 2000, the MDGs have been integrated in the planning and work agendas of many countries and international organizations. As this paper will show, many countries refer specifically to the MDGs in framing their national development goals, and an increasing number of bilateral and multilateral agencies, including the World Bank, have woven them into all aspects of their daily work. The International Monetary Fund (IMF) incorporates the MDGs in its work through its Poverty Reduction and Growth Facility (PRGF). This Discussion Paper summarizes work carried out for the 2004 High-Level Forum on Health, and reported in more depth in Foster (2004). The paper aims to draw lessons on how to accelerate progress toward the health-related MDGs, based on a literature review plus 14 country case studies of varying depth. The country cases were chosen to include all countries with completed PRSPs that were being supported by both the PRGF and a World Bank Poverty Reduction Support Credit (PRSC). Key objectives are to shed light on

• How national health goals compare to the MDGs • The content and likely cost of strategies to achieve the targets • Whether the costs can be financed, including discussion of macroeconomic

constraints on increased public expenditure • Whether increased public health expenditure can be effective, including discussion

of reforms in the management of public expenditure and of foreign assistance

Main conclusions and recommendations are highlighted at the end of each section.

2

MDGS AND NATIONAL TARGETS All of our case-study countries referred to the MDGs in setting their own national goals and targets, and many of them set national targets that are consistent with the MDGs (appendix A). Although some countries have less ambitious targets than the MDGs, all 14 target an acceleration of progress relative to the historic trend. All have targets and indicators for child or infant mortality, maternal mortality, and improved access to safe drinking water1. Nine of the 14 countries set a lower maternal mortality rate (MMR) target,2 and half of the sample countries set lower child mortality targets. All countries except Ethiopia (which starts from a very low base) have adopted more ambitious water targets, and those few with nutrition targets have also aimed beyond the MDG. All address communicable diseases, although fewer PRSPs include disease-specific targets, and there is a wider choice of targets with indicators reflecting local data availability and not necessarily in line with the MDGs. All but Nicaragua and Tajikistan address HIV/AIDS. Nutrition is universally mentioned, but few countries have specific targets, and the indicators used, again, depend on local data availability and in some cases differ from the MDGs. The MDGs do not capture all of the goals to which countries are committed and are not equally relevant everywhere (appendix B). Some countries assign higher priority to peace and stability or overall economic growth than to an exclusive focus on poverty. The specific health-related MDGs also need to be adapted to the circumstances of individual countries:

• Countries that have already met a goal may wish to set a more challenging target, as with HIV/AIDS in Uganda or access to clean water in Albania.

• The target reductions from a 1990 baseline may now be unachievable. If little progress was made, for example, in reducing child and maternal mortality in the 1990s, an even faster rate of reduction is now required to reach the 2015 targets.

• The MDG targets may be inappropriate as well as too challenging for middle-income countries like Albania where, for example, maternal mortality is already low. The MDG target of a three-quarters reduction would imply achieving levels similar to those of far wealthier countries and would involve assigning greater priority to the MDG than to other health goals (e.g., reducing chronic illness) that may be more important nationally.

• The MDGs exclude some health issues that are important in specific countries, for example, smoking in Vietnam, cervical cancer in Nicaragua. A health policy exclusively focused on the MDGs would have significant gaps.

1 Albania has near-universal access to safe water, and the target relates to household connections. 2 Ethiopia lowered a previous more challenging target.

3

• The MDGs may require disaggregation to address inequality by setting more challenging subtargets for population groups and regions with particularly poor health outcomes, like Vietnam.

SUMMARY

MDGs need to be adapted to national circumstances and priorities.

SHOULD STRATEGIES BE “NEEDS BASED,” “RESOURCE

CONSTRAINED”—OR BOTH? Though not formally linked to the MDGs, the PRSP has in practice become the main national planning instrument for articulating the strategy for achieving national goals related to the MDGs. The Millennium Project argues that national poverty reduction strategies should be “needs based,” setting out strategies consistent with reaching the MDGs and challenging the donor community to fill the financing gap left after reasonable national efforts at resource mobilization. Others stress that, to be useful, the PRSP should set out clear priorities based on a realistic assessment of available resources. By developing multiple scenarios, some countries (Rwanda, Senegal, Niger) have shown how the PRSP can be used both to guide the allocation of the resources they expect to have and as a bid for additional support. A “high” scenario is used to attract additional finance by showing what could be achieved with it, while realistic or low-case scenarios set out how priorities should be assigned to expenditure plans if fewer resources are available (WHO 2004; IMF 2003).3 The World Bank and the IMF have supported countries wishing to take this approach, but a strong case can be made for more actively encouraging it.

SUMMARY Reconcile “needs based” and “resource based” approaches by developing more than one scenario for PRSPs.

3 The “high-case” scenarios have not been directly related to achieving the MDGs, but the principle is the same.

4

FRAMING HEALTH STRATEGIES TO ACHIEVE THE TARGETS Most PRSP health strategies are dominated by health services and attempt to deliver a similar package of essential interventions that is derived from an international consensus on “what works.” The focus is on promotive and preventive interventions and on primary health care delivery. The strategies give priority to the areas most closely linked to the MDGs, with reproductive and child health and the control of the major communicable diseases given high priority. PRSPs include the costs of expanding education, roads, water and sanitation, and their contribution to achieving health targets is often recognized but seldom quantified (Country Case Studies; WHO 2004). State-funded health systems in many countries are grappling with similar problems: staff availability, pay, and motivation and the difficulty of managing a complex and geographically dispersed service with inadequate financial resources and institutional capacity relative to expectations of what can be delivered. Service coverage is low, especially in rural areas, and nearly all PRSPs envisage substantial provision for reducing geographical barriers to access, building new primary facilities, increasing operating budgets, and providing incentives for personnel to work in underserved areas. However, trends in actual health expenditure are mixed, with no strong evidence as yet of the increases in spending that would be required to implement these plans (Table 1).

5

Table 1: Trends in Public Expenditures on Health

Albania Increased health spending as share of budget and GDP since 2001, narrowing gap with Western Europe with rapid growth performance.

Benin Committed to increase health budget, but failure to disburse budgeted funds led to a reduced budget share more in line with absorptive capacity. The 2003 and 2004 budgets of roughly 2 percent of GDP represent 90 percent of the amounts envisaged in the Poverty Reduction Strategy Paper (PRSP), but actual disbursement in 2003 was only 1.5 percent of GDP—about two thirds of plan.

Burkina Faso

PRSP: Envisaged increasing the health budget share from 9.8 percent in 2000 to 11.5 percent in 2003, raising per head health spending from US$7.7 to $9.5.a

Medium-Term Expenditure Framework: Envisaged health sector share of 7.5 percent in 2004, increased to 8.5 percent in the budget due to Heavily Indebted Poor Country (HIPC) funds.

Actual spending. Just $5.90 in 2003, 6.3 percent of budget, due to low HIPC spending. Instead of increasing as envisaged, health spending has fallen as share of budget, GDP, and in real per head terms.

Cambodia Government health expenditure doubled from 0.57 percent of GDP in 1999 to 1.20 percent of GDP in 2003. Public health spending per capita increased by more than 40 percent between 2001 and 2003 but from a very low base (around $3 per head). However, domestic public expenditures represent an estimated 9 percent of total health sector expenditures

Ethiopia Spending has fluctuated as low as 5 percent of government spending since 1992/3, short of the 8.2 percent targeted in the PRSP for 2004/5. Spending of $1.50 per head in 2000/1 was possibly the world’s lowest. Regional subsidies are forecast to be flat, limiting scope for increase.

Ghana Spending (government and donors) has increased by 30 percent in real terms since 2001. Government is exceeding the 11 percent target share of recurrent budget—but due largely to increases in salaries and investment. Nonsalary recurrent budget (and productivity indicators) have fallen.

Nicaragua Spending per capita as a share of GDP increased from 2.2 percent ($17) to 2.95 percent ($22) from 2001 to 2003. Current GDP share is sufficient to meet National Development Plan cost estimates to achieve the goals.

Tajikistan Health care spending fell from 1.17 percent of GDP to 1.01 percent, and from 6.3 percent to just 5.3 percent of government spending from 2001 to 2003.

Tanzania Spending increased 75 percent in real terms in three years to fiscal 2004, but the budget share of 9.7 percent is below the Abuja 15 percent target. Spending including donors amounted to only $7.26 per capita, and the budget share dropped in both 2003 and 2004. There was some increase in the share devoted to primary and preventive care.

Uganda Health share of noninterest budget increased from 2.5 percent in 1987/88 to 9 percent in 1998/9 and 12.2 percent in 2002/3, with improved targeting. The share of Maternal and Child Health Programme (MCHP) has been raised to 15 percent target share for 2007/8. Current per head spending of between $8 and $9, will reach $11 in 2015 with15 percent share. Implementation rate has been more than 95 percent, but fell to 90 percent in fiscal 2003 due to recruitment problems.

Vietnam Due to data limitations and gaps, the trend is unclear, but a substantial increase in funding of services for the poor seems to have occurred.

a. All dollars in this report are U.S. dollars. Source: Country Case Studies.

6

Decentralization, in some form, is a nearly universal theme. Institutional approaches range from limited changes to increase the responsibility of lower level units under health ministry authority, to contractual arrangements with public and/or private entities to provide agreed levels of service, to full decentralization of responsibility to a democratically accountable lower tier of government. Efforts to help communities hold service providers accountable are also increasingly common, including community management of primary health facilities (Benin, Burkina Faso, Rwanda, Nepal). An increased community voice is often linked to community financing schemes. Pre-payment schemes increase utilization by those who are covered, but inability to pay usually excludes the poor from participation (Rwanda coverage ranges from 10 to 50 percent of population, Ghana similar). The desire to decentralize responsibility and hold those delivering services accountable for results has not always been accompanied by sufficient willingness of finance ministries, health ministries, or donors to decentralize control of the resources on which performance depends. Reasons of lack of capacity and problems of accountability are cited, but the effect is that good managers may be frustrated in their efforts to achieve improved results (Box 1).

7

Box 1: Is there a need to delegate more authority to those responsible for achieving results?

Benin. Policy in 1995 envisaged decentralization to health facilities and districts, public-private partnerships, and performance contracts. Overcentralized budget management has frustrated this policy and prevented planned increases in health expenditure but is now being addressed with PRSC support. Burkina Faso. Complex multilevel planning but overcentralized and complex procedures contribute to low budget execution (below 80 percent including Heavily Indebted Poor Country (HIPC) funds ) and very late release of funds, especially at the periphery. Key MDG priorities are heavily dependent on HIPC funds and donor projects, but there are large shortfalls in this funding. The PRSC supports limited introduction of more decentralized access to funds at district level. Ethiopia. Health sector support is inconsistent with block-grant funding of regions. Health Sector-wide Approach (SWAP) is mainly project financed, but donor projects achieve lower and more variable disbursement than government funds. Government is seeking more budget support. Ghana. The Ministry of Health (MOH) has a performance contract with the Ghana Health Service (GHS), but the MOH has retained responsibilities for procurement, staffing, and training. There are 23 administrative steps for districts to access central government funds. Nepal. Reviews in the late 1990s concluded that vertical projects were inefficient and unsustainable and urged decentralization and enhancement of the community role. Detailed budget programming and late donor confirmation of support results in late releases. But decentralization is proceeding. Sub-health posts are being handed to communities, so far with limited powers (no control over hiring staff). There are, however, plans to empower local bodies to vary compensation and to devolve drug and medical supply purchase to districts. Public-private partnerships are also being expanded. Tanzania. District health plans require formal approval by a basket funding committee. This delayed fund release in fiscal 2003. Source: Country Case Studies.

8

Households usually bear most of the costs of health care, but the consequence is low utilization of services by poor households. Most attempts to protect the poor’s access to services by exempting them from payment have failed. Abolition of user fees can lead to big increases in utilization (Uganda), but poses risks to quality, and the increased demand may not be sustained unless there is an increase in facility-level operating budgets to replace the lost revenue. Several case study countries are piloting approaches to reducing or eliminating cost barriers to the poor. PRSPs say little about how nonpriority expenditures can be contained to permit expanded services for the poor, with Nicaragua one of the few making explicit reference to increased reliance on the private sector to fund services for the better-off. In Ghana, the stated priorities (district services and nonsalary recurrent spending) have been squeezed by big increases in spending on salaries, on investment, and on central spending; in Burkina Faso, the primary health share of the total health budget has fallen, whereas policy is to increase it.

SUMMARY • Stronger interdepartmental coordination is needed to move from “health services strategy” to

“health strategy.” • Plans for expanding support to high-priority interventions (on which there is broad agreement)

need to be balanced by plans for how funding can be withdrawn from lower-priority services. • In most countries, achieving national targets requires institutional reforms to strengthen

performance incentives, but with accountability for results matched to more reliable access to the necessary resources.

ESTIMATING THE COSTS OF ACHIEVING THE TARGETS Health is just one of the sectors competing for scarce funds to accelerate progress toward national goals and the global MDGs. Explicit assumptions linking public expenditure costs to expected health outcomes are necessary to make the case for increased resources, both with the Ministry of Finance and with the donors. They can also help set budget priorities in light of available resources, and provide a quantified framework for subsequent monitoring and evaluation, and for setting realistic targets for managing performance. The shortfall in resources has been highlighted in global estimates, but most national poverty reduction strategies do not estimate the public expenditure cost of achieving their targets (Soucat et al. 2002). Most of the health strategies we looked at have been costed, but few of them articulate clearly (and provide evidence for) the assumed chain of causality linking the resources required to the activities to be undertaken, the outputs to be produced, and the expected impact on outcomes. Many countries still produce incremental health budgets not linked to objectives or even activities (e.g., Tajikistan). Others produce detailed activity-based budgets in which it is difficult to link the many objectives and activities to overall strategic priorities (Ghana, Tanzania). In most cases,

9

the goals are determined with reference to the MDGs and to the expectations of the donors, the plans reflect national constraints and priorities and available resources, but the links between the two are not specified. A more explicit treatment of the links from resources to outcomes would probably reveal that the health goals are in most cases inconsistent with health budgets. None of the countries is close to meeting the 15 percent health sector budget share agreed by African countries in Abuja, more than half of them are spending less than $10 per head on health, yet in several countries the share is both low and not increasing (Ethiopia, Benin, Burkina Faso, Tajikistan, and Tanzania in the 2004/5 budget). By accepting targets that may be inconsistent with the proposed resources, health ministries risk undermining public confidence and the morale of health workers, while missing an opportunity to make a convincing case for an increased budget. Different approaches to costing are in use for different purposes. Estimates of the cost-effectiveness of specific health interventions have provided a sound basis for deciding what should be included in essential services packages. For costing national strategies and relating them to outputs, however, it is necessary to consider the cost and impact of measures to address institutional and incentives problems and to prepare cost estimates that can be “mapped” to the way that budgets are actually allocated. One approach that is being piloted aims to do this by packaging interventions in terms of how they are delivered (facility-based, outreach, community-based) and focuses on bottlenecks constraining coverage and effectiveness (physical accessibility, human resources, logistics and supply, cost and other barriers to demand and utilization, gaps in technical and organizational quality, and steering and management costs) (Soucat et al. 2002). The approach can help make the case for increased funding by identifying the areas affording the most scope for significant impact on health outcomes at modest cost. In Mauritania, the 40 percent increase in the health budget in 2002 was reportedly influenced by analysis suggesting that a targeted increase could achieve within five years a 30 percent reduction in child mortality and a 40 percent reduction in maternal mortality (Soucat et al. 2002).4. The approach can also reveal how marginal costs rise as the MDG targets are approached. In Ethiopia, a 42 percent reduction in under-5 mortality could be achieved for less than half the cost required to achieve the 66 percent reduction called for to achieve the MDG. Though useful as a conceptual approach, the realism of the analysis has yet to be confirmed in terms of actual outcomes.

SUMMARY

• Strategies should include explicit analysis of expected linkages between costs-outputs-outcomes.

• Cost estimates should address institutional constraints and be prepared in a form that can be mapped to budgets and support resource bids.

• Priorities should be identified to permit adjustments in the light of resource availability.

4 Knippenberg et al (2003) report that application of the marginal budgeting for bottlenecks approach for preparing health MTEFs in Mali and Mauritania resulted in doubling of health budgets.

10

ARE MACROECONOMIC FRAMEWORKS TOO RESTRICTIVE? There is some debate over the wisdom of increasing public expenditure as steeply as would be required to reach the MDGs, even if such increases could be financed from external grants. Aid only transfers real resources when net imports increase. If the aid is spent on domestic goods and services instead of imports, then the non-aid sector of the economy must be persuaded to absorb the aid by exporting less and importing more. This increases the availability of domestic goods and services to meet the aid-financed demand for them. If the increased expenditure financed by aid does not result in increased net imports, it will simply squeeze out existing customers for the goods and services bought with the aid, an effect that could be equally well achieved without aid by using fiscal and monetary policy. The problem of “Dutch disease” arises because the increase in demand for net imports required to absorb the aid may require some appreciation of the real exchange rate, potentially damaging the competitiveness of the traded goods sectors that are thought to be of particular importance for economic growth5.

These problems of aid absorption are very relevant to the health sector, where local costs typically account for between 70 and 75 percent of total public sector health spending (Millennium Project (2003). As described in the previous paragraph, increased public expenditure on health therefore implies a re-orientation of domestic resources towards producing more public sector health services, and fewer private sector goods and services, particularly traded goods and services. This may be desirable if the additional public sector health outputs are more socially valuable than the private sector outputs they displace. That may well be the case with cost-effective health spending, especially when the positive impact of improved health on productivity is taken into account alongside the social benefits. However, if we assume that the public sector expenditures with the highest benefit: cost ratios are undertaken first, while the least profitable private sector activities are displaced first, the diminishing marginal benefits of additional public expenditure will eventually fall below the rising marginal costs of displacing private sector activity. Irrespective of Dutch disease, there will come a point beyond which increased public expenditure on local costs should not be undertaken, even if financed by grants. The argument is not about whether such limits are needed but concerns the judgment on where they should be set.

These issues are not discussed in most PRSPs (IEO 2004), perhaps because governments recognize that the macroeconomic framework has to be negotiated with the IMF, because the existence of an on-track IMF program remains a prerequisite for accessing significant external aid or HIPC debt relief.

5 IMF (2005), The Macroeconomics of Managing Increased Aid Inflows: Experiences of Low-Income Countries and Policy Implications. Policy Development and Review Department, August 8, 2005.

11

The IMF has often been accused of preventing countries from adopting ambitious spending plans, even when financed from aid. However, IMF policy is to support increased aid, stating that additional aid inflows should be accommodated by appropriate adjustments of the program’s fiscal and financing targets, “if they can be effectively absorbed and utilized without endangering macro stability.” (IMF 2004). Moreover, IMF programs have actually overestimated foreign aid, overestimated GDP growth, and consequently overestimated both domestic and foreign resources available to finance public expenditure (IMF and IDA 2003).6 Despite this evidence, it could still be argued that the IMF may unintentionally restrain future aid commitments by negotiating fiscal frameworks that assume only modest growth in aid. Countries may not push for additional aid flows, nor will donors offer such aid, if the macroeconomic projections on which the expenditure program is based do not show a clear need for it. Table 2, showing the assumptions for our sample countries, does suggest a conservative bias in the projections, all but one of which converge toward a level where public expenditure as a share of GDP is around 25 percent. No country is projected to spend more than 30 percent of GDP on public expenditure. Countries spending more than 25 percent of GDP assume a flat or declining share of public spending. Only two countries assume more than a 2 percent of GDP increase in the share of public expenditure within the projection period, both of them countries starting from a low base of public expenditure (less than 20 percent of GDP). With per capita economic growth typically forecast at between 3 and 4 percent a year, the projections imply an increase of about 50 percent in public expenditure per capita by 2015. This may sound like a lot, but it would leave all but three of our countries spending less than $20 of public funds on health—still far short of estimates of the cost of delivering the essential health package to all. In a 3 to 7 year projection period, less than half of our sample countries project any increase in net external financing as a share of GDP. Even relatively well-performing and low-income countries are projecting aid increases for themselves that are well below the commitments made at Monterrey. The fiscal frameworks in IMF programs are difficult to assess, because the Fund provides no clear justification for the assumptions on the level and financing of public expenditure (IEO 2003). The convergence of spending toward roughly 25 percent of GDP is not a result of any IMF policy, although it may reflect a tendency for IMF country staff to encourage countries to move toward a level of expenditure that Board and senior management have found acceptable in other cases. Whatever the explanation, the lack of variation is surprising. Countries differ in their needs: low-income countries with poor infrastructure and low education and health levels could make a strong case for a higher public expenditure share to create the conditions for faster economic growth and poverty reduction7. Countries differ in their ability to finance expenditure, both their ability to

6 This overestimation of the available resources is just as serious as underestimation, because unplanned shortfalls in resources may necessitate damaging short-term cuts in public spending, often focused on easy to cut nonsalary recurrent budgets. Overestimation thus results in disproportionate negative impacts on outputs by denying staff members the resources to do an effective job. 7 See for example, Millennium Project (2004).

12

attract external aid and the scope for mobilizing domestic resources without damaging growth and stability. It would also be reasonable to expect that, other things being equal, public expenditure would be higher in countries like Uganda and Burkina Faso, where almost all aid flows are recorded in the IMF public expenditure tables, than in countries like Benin, where more than half of aid is not included (Tables 3–5). The lack of explicit rationale for the assumptions, together with the absence of the expected degree of difference in projected spending between countries, adds up to a strong case for a more open debate on the macro framework. The PRSP preparation and annual review process may provide an appropriate forum for the wider debate on the level and means of financing public expenditure. Both Ministry of Finance and IMF views need to be reflected in the PRSP. Good practice in this regard includes:

• The government personnel who prepare the macroeconomic framework for the budget should be responsible for (or at least heavily involved in) developing the fiscal envelope for the PRSP.

• The IMF should also be involved at an early stage, through the resident representative.

• Open debate supported by high quality technical analysis. In Uganda and Tanzania, independent macroeconomic analysis that commanded the respect of IMF staff was helpful in persuading the Fund to accommodate higher expenditure.

• The PRSP needs to be adapted in the light of events. Several countries have coordinated the annual PRS review with adjustments to the targets, the macroeconomic framework, and the expenditure priorities and budget ceilings.

Another problem is that aid commitments are short-term and unreliable, whereas the additional public expenditure that is needed is mainly for recurring costs that will need to be sustained and to grow into the indefinite future. The problem is further reinforced if a temporary surge in aid causes Dutch disease, damaging the potential of the economy to finance higher spending through economic growth. Countries therefore face significant risks if they establish health systems that cannot be maintained if donor preferences change.8 A number of approaches could be taken to managing this problem: the International Financing Facility as a way to ensure growing aid at least at global level; increased reliance on multilateral channels less subject to political pressures; longer-term commitments to specific expenditure programs, with guarantees that they will continue as long as the program-specific conditions are met; compensatory mechanisms to smooth fluctuations caused by shortfalls by individual donors; further debt relief as a form of irrevocable long-term commitment; increased use of aid for reserve build-up to help manage aid fluctuations. Each of these requires donor willingness to commit their money longer term in ways that are less at risk of interruption by events outside program objectives.

8For poor countries such as Tanzania and Ethiopia, sustaining per capita expenditure increases financed by a doubling of aid flows would require the higher level of aid to be maintained for 20 years or more. See Foster (2003).

13

SUMMARY

• Involve the MOF and IMF in preparing PRSP macro-fiscal frameworks to ensure consistency with the budget, but with explicit rationale and more open debate.

• Coordinate annual PRSP progress review with budget cycle including revision of macro-fiscal framework and budget ceilings and priorities.

• Reduce risks of aid dependence by longer-term commitments less prone to interruption, more reliable timing of disbursements.

14

Table 2: IMF Macroeconomic Frameworks in Case-Study Countries

Country GDP growth per year

(geometric average, percent) Domestic revenue

(percent GDP)

Net external financea of public expenditure

(percent GDP)

Public expenditure,

(percent GDP)

Historical

(years)

Projected

(years)

Base

(year)

Projected

(year)

Base

(year)

Projected

(year)

Base

(year)

Projected

(year)

Albania 6.4

(1999–2003)

6.1

(2003/7)

22.0

(2003)

22.4

(2007)

1.7

(2003)

1.9

(2007)

26.5

(2003)

26.6

(2007)

Benin 5.6

(1999/2003) 6.6

(2003/6)

16.8

(2003)

17.0

(2006)

4.8

(2003)

3.5

(2006)

21.0

(2003)

22.0

(2006)

Burkina Faso 6.6

(1999–2003)

5.4

(2003/6)

12.4

(2003)

14.6

(2006)

8.7

(2003)

5.7

(2006)

21.6

(2003)

23.4

(2006)

Cambodia 5.8

(1999–2003)

4.7

(2003/9)

10.4

(2003)

14.0

(2009)

6.0

(2003)

23.4

(2005)

21.1

(2009)

17.4

(2003)

18.4

(2009)

Ethiopia 3.7b

(1999–2003)

6.8

(2003/6)

19.6

(2003)

20.4

(2006)

14.8

(2003)

8.6

(2006)

29.1c

(2001)

29.5

(2006)

Ghana 4.7

(1999–2003)

5.0

(2003/8)

20.8

(2003)

22.4

(2008)

8.0

(2003)

5.1

(2008)

29.0

(2003)

24.3 d

(2008)

Nepal 2.2

(2000/3)

4.5

(2003/6)

12.3

(2003)

13.5

(2006)

2.9

(2003)

4.2

(2006)

16.3

(2003)

18.2

(2006)

Nicaragua 2.6

(1999–2003)

4.2

(2003/8)

21.9e

(2003)

22.1

(2008)

10.5

(2003)

8.7f

(2004)

30.3

(2003)

28.7

(2004)

26.8g

(2008)

Rwanda 5.6

(2000/3)

5.3

(2003/6)

13.5

(2003)

13.6

(2006)

10.5

(2003)

17.1

(2004)

13.9

(2006)

24.1

(2003)

28.3

(2004)

25.1

(2006)

Tajikistan 9.5

(1999–2003)

5.9

( 2003/10)

16.8

(2003)

19.1

(2010)

2.8h

2003)

1.2

(2010)

19.0

(2003)

22.1

(2010)

Tanzania 7.6

(1999–2003)

6.1

(2003/6)

11.4

(2003)

14.0

(2007)

7.7

(2003)

10.7

(2007)

18.6

(2003)

25.4

(2007)

Uganda 5.6

(2000/3)

5.8

(2003/8)

12.3

(2003)

13.4

(2005)

11.5

(2003)

8.5

(2005)

23.7

(2003)

22.9

(2005)

Vietnam 5.4

(1999–2002)

6.8

(2002/7)

22.5

(2002)

22.2

(2007)

1.4

(2002)

1.6h

(2004)

24.8

(2002)

25.1

(2007)

Note: Nepal, Rwanda and Vietnam numbers are based on 2003 PRGF program. The other countries are based on latest 2004 reported programs. a. Including grants. b. GDP (at factor cost) growth rate at constant prices. c. Includes special programs such as demobilization and reconstruction, which accounts for around 0.7 percent of GDP. d. Interest payment as percent of GDP decreases from 6.2 percent in 2003 to 1.7 percent in 2008. e. These numbers are reported as total current revenue as percent of GDP, capital revenue is recorded under the capital expenditure (net of capital revenue). f. Due to missing data for some of the financing projections, projection target year is different (2004) for net external financing. IMF Executive Board Completes Fifth and Sixth Reviews under Nicaragua's PRGF Arrangement (Sept. 2004). g. Interest payment as percent of GDP decreases from 5.1 percent in 2003 to 2 percent in 2008. h. These are data for net foreign borrowing. Grants are excluded due to missing projections for the program years (2005–2010). The program target year is different due to missing data for some of the financial projections.

15

COORDINATING HEALTH PLANS WITH THE MTEF & THE BUDGET

The public expenditures implied by the PRS need to be implemented via the national budget. Good-practice approaches include:

• The PRSP establishes clear priorities and criteria, reflected in the guidelines and ceilings given to line ministries for preparing budget and MTEF proposals.

• Budget allocations are the outcome of an iterative process in which proposals by line ministries are scrutinized by the center, and adjusted in the light of national priorities and resources.

• There is an annual process for reviewing sector-level physical and financial progress, and the domestic and foreign financing requirements for the coming period, involving government and development partners, and timed to inform budget preparation.

• The incentives for line ministries to review their performance, construct well-designed budgets that shift resources toward national goals, and present them in ways that make the strategic shifts transparent, are reinforced by central scrutiny of budget proposals, challenging ministries to justify their bids.

• The Ministry of Finance and Cabinet establish and maintain the credibility of the process by ensuring that carefully prepared budgets that reflect national goals, receive favorable treatment in the budget that is finally approved, and in the timely and full release of funds.

• The approved MTEF is the same as the annual budget for the first year, and the chart of accounts is structured in such a way that spending priorities of particular importance for achieving the MDGs can be identified.

• The Ministry of Finance provides credible medium-term assurances of sectoral budget levels or shares to encourage line ministries to reallocate resources from lower priority areas without fearing that their budgets will suffer. Credibility can be built via a medium-term track record in which it is shown by example that the MTEF guides resources. Agreements with external partners on the health share of spending can also help build MOH confidence.

Many of these good-practice features are present in Albania, Benin, Rwanda, Tanzania, and Uganda, although the identification of priority expenditure programs is in some cases limited to the aggregate level (e.g., sector or subsector shares such as primary health), and only Uganda has a strong central budget “challenge” function. In countries with decentralized budget responsibility such as Vietnam and Ethiopia, it may be impossible for government to establish centrally a medium-term framework to determine public expenditure shares. The goal of achieving similar shifts in priority is being addressed

16

through increased resources for targeted national programs such as the province-level poverty health funds in Vietnam. Not all MTEFs have focused sufficiently on achieving a strategic shift in expenditures toward national priorities. The MTEFs in Cambodia and Ghana, and to some extent in Tanzania, are based on detailed bottom-up activity costing, resulting in bulky documents where it is difficult or impossible to see how the changes in the budget allocations relate to higher-level goals and targets. Where budget preparation and public expenditure management are particularly weak, there may be no effective means to ensure that health strategies are implemented. The effort devoted to preparing a health strategy and a sector MTEF is largely wasted if the annual budget is not in practice implemented and the medium-term priorities are not respected. Several of the countries in our sample still lack any credible mechanism for linking policy priorities to expenditure allocations, with budgets still prepared on an incremental basis, via fragmented parallel processes, and with actual budget execution not reflecting the approved budgets. In Tajikistan, for example, the budget is prepared incrementally on a line item basis; actual expenditure bears little relation to approved budgets, making it impossible to relate health outputs to budgets either for planning or reporting purposes. Cambodia is an extreme example, with health centers receiving less than 10 percent of their budgets, but several other countries need to establish credibility. Improvements in public expenditure management require action by central economic ministries, although implementation may also require reform and capacity building at sector and local government levels.

SUMMARY • Minimum standards of public expenditure management need to be attained before

any health strategy can be effective. • In good-practice cases, PRSP identifies spending priorities in consultation with

sectors, MTEF-budget process shifts resources toward them, and priorities are reviewed and adjusted each year in light of performance.

ABSORPTIVE CAPACITY Very large increases in funding over a very short period of time, as envisaged in some estimates of the cost of reaching the MDGs, might well lead to difficulties in making good use of the money. However, in most of our case-study countries funding is inadequate at present and is increasing at a speed that could be effectively used if appropriately prioritized and managed without excessive bureaucracy. For example, existing capacity is underused in many countries, because low operating budgets mean that medicines and other consumables are unavailable or have to be paid for, while personnel lack the travel budget to expand outreach. Increased aid for nonsalary

17

recurrent costs and for financing free basic health services would enable use of that existing capacity. Reducing user fees can increase the utilization of services, but only if quality does not decline, which requires the lost revenue from fees to be replaced by increased budget funding. Increased funding would also help overcome capacity problems caused by staff vacancies and by low output from underpaid personnel who need to devote time to alternative occupations or private practice. These problems can be relieved with additional funding for recruitment and for salary increases, although the increase would need to be sustainable and should preferably be sequenced as part of an approach that links increased pay to improved performance. In time, contractual arrangements with non-government agents could be put in place to expand services and support capacity development A balanced increase in funding that addresses the critical constraints in a logical sequence could be well used in most of the countries considered. However, very large disease-specific programs (such as proposed HIV/AIDS treatment programs in Guyana and Tanzania that envisage spending sums equal to half of the existing health budget) may both experience and cause capacity problems by drawing disproportionately on available staffing and other resources. Concerns about absorptive capacity frequently reflect concerns about governance and accountability rather than technical limits on spending, and procedural requirements intended to address those concerns cause the disbursement problems. In some cases, governance and expenditure management constraints are so pervasive that major reforms need to precede or accompany increased funding (Tajikistan, Cambodia). In other cases, government procedures are overcentralized and bureaucratic and need to be reformed to permit available funding to be spent (Benin, Burkina Faso). Donor project or pooled funding procedures are often part of the problem (Box 1). Donor procedures not only cause low disbursement but also divert capacity away from service delivery toward servicing the donor demand for meetings, field trips, reports, accounts, audits, and so on. By absorbing the capacity of financial management personnel, they impede government action to address the weaknesses that make parallel procedures necessary.

SUMMARY • On present trends, the binding constraint is lack of finance, not lack of capacity. • Capacity problems can be managed if health strategies tackle bottlenecks in a

logical sequence, and avoid large “earmarked” commitments that distort health sector priorities.

• Where government is committed to improving financial management, external partners should use government systems while supporting coordinated action to strengthen them as necessary.

18

REFORMING DEVELOPMENT ASSISTANCE Development agency procedures are major constraints on the execution of public expenditure plans. Commitments are too short term to be the basis for long-term strategies. Disbursements often fall long short of commitments and are unpredictable and subject to interruption for reasons outside the program itself. Most development assistance is still committed through parallel arrangements and is imperfectly coordinated with government strategy. Box 2, based on Appendix C, presents some rough estimates of the significance of aid to the budget. Budget support is increasing as a share of donor support, but even in the most highly aid-dependent countries it represents little more than half of the support for public expenditure and less than one third of total aid flows. On average, less than 20 percent of donor disbursement is provided as budget support. This is a major problem in aid-dependent countries where more than half of health spending may be donor funded and where the number of donors involved continues to increase. The priorities for increased spending are dominated by recurrent costs, spent via geographically scattered cost centers and difficult to support via projects without incurring high transaction costs. If the case is accepted for an increase in health spending in order to progress toward the MDGs, it seems inescapable that the bulk of the increase will need to be provided as budget support.

19

Box 2: Where does all the aid go?

On average, for every $1 disbursed by donors to our 14 case study countries, we estimate: Direct donor spending (technical assistance and direct payments) not recorded in balance of payments

$0.30

Recorded in balance of payments, but not reported as part of government spending

$0.20

Aid earmarked to specific projects $0.30Provided as budget support $0.20

Some of the budget support is itself earmarked to specific sectors or budget lines. Source: Appendix C. Meanwhile, countries continue to face the problem of coordinating large numbers of donors providing their assistance via multiple routes. Good practice approaches from our sample countries include:

• Joint Government-donor reviews of sector performance that are coordinated with the MTEF and national budget process and will feed in to a national PRS progress report or into a national public expenditure review process. There are a number of country examples where arrangements along these lines are in place, notably Uganda and Tanzania, where external partners work closely with government in sector working groups, and where the sector dialogue is coordinated around the annual budget cycle.

• Indications of future donor funding should be made early enough in the year to be taken into account in setting ceilings for budget preparation, and should be confirmed as the budget is being finalized.

• Donor policy dialogue at different levels needs to be coordinated. The PRSC and general budget support groups are the appropriate vehicles for addressing issues that are cross-cutting or are the concern of the central economic ministries. Of direct relevance to health, this includes the overall macro-economic framework, budget allocation, public finance management, civil service reform, and decentralisation. Where there is an established sector dialogue, the health content of the PRSC should rely upon the sector reviews to set and assess the achievement of sector level actions, as is the case in Uganda, avoiding overloading the budget support policy matrix with sector level detail.

20

Several of the country case studies suggest that large commitments from the global funds are distorting priorities in a number of countries: committing an unsustainable share of the budget to HAART for AIDS sufferers, sucking staff and resources into vertical programs with costs that are neither replicable nor sustainable without longer-term commitments than the donors have yet provided (Tanzania, Guyana, Ethiopia). GFATM has in some countries set up separate coordination arrangements specific to the funds it is providing9. Several countries are uncomfortable with the approach of some of these new actors. In August 2004, the Ministry of Finance in Uganda was reported as having decided to cap new project aid commitments that are outside the national health strategy10. There is a strong case for all external partners in the health sector working entirely within existing health sector coordination arrangements, and providing their assistance in support of the PRSP strategy, focusing first on filling the financing gaps for implementing the PRSP.

SUMMARY • Progress toward the MDGs requires a further shift toward budget support as the

main aid modality in aid-dependent countries. • All donors should participate in sector coordination and provide sufficient

information for government to include their commitments and disbursements in the macroeconomic framework and reflected in public expenditure plans.

• Where government has a sound sector strategy, the first call on donor funds should be to ensure that it is fully funded.

• Donors should try to commit early enough to inform the budget preparation. • Where a strong sector level policy dialogue is in place, the PRSC should rely on

sector reviews to agree sector level actions and to assess their achievement.

9 Barbara Bruns, informal note of June 28, 2004 workshop on the role of the IMF in relation to HIV programs. 10 New Vision, Uganda, August 20, 2004.

21

APPENDIX A: GOALS, TARGETS, AND INDICATORS FOR MDGS 1 THROUGH 7

GOALS AND TARGETS

INDICATORS GOAL 1: ERADICATE EXTREME POVERTY AND HUNGER

Halve, between 1990 and 2015, the proportion of people whose income is less than US$1 a day

1. Proportion of population below $1 (PPP) per day 2. Poverty gap ratio [incidence x depth of poverty] 3. Share of poorest quintile in national consumption

Halve, between 1990 and 2015, the proportion of people who suffer from hunger

4. Prevalence of underweight children under-five years of age 5. Proportion of population below minimum level of dietary energy

consumption

GOAL 2: ACHIEVE UNIVERSAL PRIMARY EDUCATION

Ensure that, by 2015, children everywhere, boys and girls alike, will be able to complete a full course of primary schooling

6. Net enrolment ratio in primary education 7. Proportion of pupils starting grade 1 who reach grade 5 8. Literacy rate of 15 to 24 year-olds

GOAL 3: PROMOTE GENDER EQUALITY AND EMPOWER WOMEN

Eliminate gender disparity in primary and secondary education preferably by 2005 and to all levels of education no later than 2015

9. Ratios of girls to boys in primary, secondary, and tertiary education 10. Ratio of literate females to males of 15 to 24 years of age 11. Share of women in wage employment in the nonagricultural sector 12. Proportion of seats held by women in national parliament

GOAL 4: REDUCE CHILD MORTALITY

Reduce by two thirds, between 1990 and 2015, the under-five mortality rate

13. Under-five mortality rate 14. Infant mortality rate 15. Proportion of 1 year-old children immunized against measles

GOAL 5: IMPROVE MATERNAL HEALTH

Reduce by three quarters, between 1990 and 2015, the maternal mortality ratio

16. Maternal mortality ratio 17. Proportion of births attended by skilled health personnel

GOAL 6: COMBAT HIV/AIDS, MALARIA AND OTHER DISEASES

Have halted by 2015 and begun to reverse the spread of HIV/AIDS

18. HIV prevalence among 15 to 24 year- old pregnant women 19. Condom use rate of the contraceptive prevalence rate 20. Number of children orphaned by HIV/AIDS

Have halted by 2015 and begun to reverse the incidence of malaria and other major diseases

21. Prevalence and death rates associated with malaria 22. Proportion of population in malaria risk areas using effective malaria

prevention and treatment measures1 23. Prevalence and death rates associated with tuberculosis 24. Proportion of tuberculosis cases detected and cured under directly

observed treatment short course (DOTS) GOAL 7: ENSURE ENVIRONMENTAL SUSTAINABILITY

Integrate the principles of sustainable development into country policies and programs and reverse the loss of environmental resources

25. Proportion of land area covered by forest 26. Ratio of area protected to maintain biological diversity to surface area 27. Energy use (kg oil equivalent) per $1 GDP (PPP) 28. Carbon dioxide emissions (per capita) and consumption of ozone-

depleting CFCs (ODP tons) 29. Proportion of population using solid fuels

Halve, by 2015, the proportion of people without sustainable access to safe drinking water

30. Proportion of population with sustainable access to an improved water source, urban and rural

By 2020, to have achieved a significant improvement in the lives of at least 100 million slum dwellers

31. Proportion of urban population with access to improved sanitation 32. Proportion of households with access to secure tenure (owned or rented)

22

APPENDIX B: MDGS AND NATIONAL GOALS IN CASE-STUDY COUNTRIES

This appendix comes largely from the World Bank data base, judgments on degree of ambition based on whether the implied rate of progress is consistent with that needed to reach the target. In many cases, progress was slow or negative for the first decade after 1990, which would require catch-up to reach the MDG; hence some of my judgments differ from those of the World Bank, noted in footnote.

Infant Mortality

Country Current status MDG Target 2015 PRSP/national target Comparison of national target to MDG

Albania 21 (2001) 9 17 (2004) 10 (2015)

Marginally less ambitiousa

Benin 93(2002) 37 65 (2005) 28 (2015)

More ambitious

Burkina Faso

105 (2000) Two-thirds reduction on 1990, baseline unclear

70 (2005) 45 (2014)

More ambitious

Cambodia 95 (2000) Two-thirds reduction on 1990, baseline unclear

90 (2005) Less ambitious

Ethiopia 97 (2001) Two-thirds reduction on 1990, baseline unclear

85 (2005) Less ambitious

Ghana 56 (2000) Two-thirds reduction on 1990, baseline unclear

50 (2005) Less ambitious

Guyana IMR 58 (1999) 83% Measles immunization

IMR 25; 100% measles immunization

IMR 42; 92% measles immunization (2005)

More ambitious

Nepal 64 (2001) Two-thirds reduction on 1990, baseline unclear

45 (2006) More ambitious

Nicaragua IMR 40 (1998) Two-thirds reduction on 1990, baseline unclear

IMR 32 (2005) Less ambitious reduction to 2005, but 2015 target is accepted.

Rwanda 107 Two-thirds reduction on 1990, baseline unclear

35 (2015)

Tajikistan 36.7 (2001) Two-thirds reduction on 1990, baseline unclear

32 (2006) Less ambitious

Tanzania IMR 99 (2000) 33 50 (2010); 20 (2025) Less ambitiousb Uganda 88 (2000) 31 (2015) 68 (2005) Consistent with MDG Vietnam Two-thirds reduction on

1990, baseline unclear IMR 30 (2005), <25 (2010)

23

Child Mortality

Country Current status MDG Target 2015 PRSP/national target Comparison of national target to MDG

Albania 14 Benin 151 (2002) 61 117 (2006) More ambitious Burkina Faso

69

Cambodia 125 (2001) 38 118 (2005) Less ambitious Ethiopia 167 (2001) 64 160 (2005) Less ambitious Ghana 110 (2000) 42 95 (2005) Less ambitious Guyana 30 Nepal 91 (2001) 48 72 (2006) More ambitious Nicaragua CMR 50 (1998) 22 CMR 37 (2005); targets for

immunization and reducing incidence of diarrhea, ARI .

Less ambitious to 2005, but 2015 target is accepted

Rwanda 107 (2000) 59 35 (2015) More ambitious Tajikistan 42 Tanzania 161 (1999/2000) 48 79 (2010) Less ambitious Uganda 160 (1995) 54 103 (2005) Consistent with MDG Vietnam 38 (2001) 17 <5MR 36 (2005), <32 (2010) Less ambitious

Maternal Mortality

Country Current status MDG Target 2015 PRSP/national target Comparison of national target to MDG

Albania MMR 20 (2000) 10 15 (2004) 11 (2015)

Marginally less ambitiousc

Benin 495(Govt) 850 (WB/UN)

248 (WB/UN) 475 (2006) 390 (2015)

Less ambitious

Burkina Faso

484 (2000) Three-quarters reduction on 1990, baseline unclear

350 (2004) 200 (2014)

Rate of decline more ambitious

Cambodia 437 (1998) Three-quarters reduction on 1990, baseline unclear

372 (2005) Less ambitious

Ethiopia 600, revised to 500 in progress report (2001)

Three-quarters reduction on 1990, baseline unclear

425, revised to 450 Target revised from more to less ambitious

Ghana 200 (2000) Three-quarters reduction on 1990, baseline unclear

160 (2005) Less ambitious

Guyana 190 (1997) 69 130 (2005) Less ambitious Nepal 415 (2001) Three-quarters

reduction on 1990, baseline unclear

300 (2006) More ambitious

Nicaragua MMR 148 (1999) Three-quarters reduction on 1990, baseline unclear

MMR 129 (2005) Less ambitious to 2005, but 2015 target is accepted

Rwanda 810 (2000), qualified attendance 35%

Three-quarters reduction on 1990, baseline unclear

MMR 202(2015) Qualified attendance 60% (2010)

More ambitious rate of decline

Tajikistan 43 (2001) Three-quarters reduction on 1990, baseline unclear

40 (2006) Less ambitious rate of decline

Tanzania 529 (1996) MMR 133; 90% attended births.

450 (2003), 265 (2010); 80% attended births (2010)

Less ambitious

Uganda 505 (2000) 131 354 (2005) Consistent with MDG Vietnam 160 (1995) Three-quarters

reduction on 1990, baseline unclear

80 (2005), 70(2010) Less ambitiousd

24

HIV/AIDS

Country Current status MDG Target 2015 PRSP/national target

Comparison of national target to MDG

Albania Less than 100 known cases 2002

Halt and reverse by 2015 Prevent HIV/AIDS taking hold

MDG target not appropriate.

Benin 4.1% (pregnant women)

Halt and reverse by 2015 <6% Not comparable

Guyana 237 reported cases, 74 women 15 to 45 yrs (1999)

Halt and reverse by 2015 161 reported cases (64 women 15 to 45 yrs) in 2015

More ambitious

Tanzania 7.8% 2001 <8% Reduce prevalence and promote behavior change

More ambitious (MDG achieved)

Uganda 6.5% Rate peaked 1992/3.

Halt and reverse by 2015 5% (2005) More ambitious

Vietnam Indicators but no target Other Infectious Diseases

Country Current status MDG Target 2015 PRSP/national target

Comparison of national target to MDG

Benin Reduce malaria death rate below 1990 7.6/100,000

6.2/100,000 by 2015 More ambitious

Tanzania Malaria case fatality rate Vietnam Indicators but no target

Water and Sanitation (% access to safe water unless otherwise indicated)

Country Current status MDG Target 2015 PRSP/national target

Comparison of national target to MDG

Albania 97% 99% Benin 66% (2001) Halve, by 2015, the proportion of

people without sustainable access to safe drinking water

74% (2005) More ambitious

Cambodia 29% (1999) Halve, by 2015, the proportion of people without sustainable access to safe drinking water

40% (2005) Consistent

Ethiopia 30% (2000) 62% 39% (2005) Less ambitious (no improvement in 1990s)

Ghana 76% Guyana 92% (1999) Halve, by 2015, the proportion of

people without sustainable access to safe drinking water

95% 2003,100% urban by 2005

More ambitious rate of change

Nepal 72 (2001) 83% 85 (2006) More ambitious Nicaragua

Water 67%, sanitation 84%

85% Water 75%(2005), sanitation 88% (2004)

More ambitious

Rwanda Safe water 52%, sanitation 15%

Halve, by 2015, the proportion of people without sustainable access to safe drinking water

Safe water 72%, sanitation 56% (2015)

Tajikistan 51% (2001) Halve, by 2015, the proportion of people without sustainable access to safe drinking water

58% (2006) More ambitious rate of change

Tanzania 58% national 48% rural

79%e 85% rural (2010) More ambitious

Uganda 55% (2002/3) 72% 100% (2015) More ambitious Vietnam 77% (2000) 77% 80%urban, 60%rural have

safe water by 2005; 85% rural by 2010.

More ambitious (MDG target achieved)

25

Education and Gender Equality

Country Current status MDG Target 2015 PRSP/national target Comparison of national target to MDG

Albania NER 93% (2002); girl: boys 1.02

100% 100% Same; gender equality already achieved.

Benin NER 81% 84% complete; 0.65 girls/boy enrolled

100% complete primary by 2015; Girl/boy ratio 1.0 by 2005

99% net enrollment rate (NER) and 1.0 girl: boy ratio by 2015.

Marginally less ambitious

Burkina Faso

41% NER 100% 70% “in 10 years”; no gender target given.

Less ambitious

Cambodia NER 84% (2001) 100% Guyana NER 97% (1999) 100% NER and primary

completion rate (PCR) 100% NER 2005 More ambitious

Nepal NER 80 (2001) 100% NER 90 (2006) More ambitious Nicaragua NER 75% (1999) 100% 83% NER (2005), no gender

target. Consistent growth ratef

Rwanda NER 72 (2000) NER 100 (2015) Same Tajikistan NER 100 (2015) NER 90% (2015) Less ambitious Tanzania NER 60%

1999/2000 100% NER 100% literacy 2010, NER

70% 2004; gender equality in enrolment 2005.

More ambitious growth rate

Uganda NER 83-90% (poorest to richest), P7 year NER 10% (2002); girl: boy 0.98

100%, girl: boy 1.0 100% (2003), 1.0 girl: boy in primary

More ambitious date for universal primary education (UPE). No secondary gender target.

Vietnam 95% NER (2000) 100% NER NER97% 2005,99% 2010; gender gaps closed 2005, ethnic gaps closed 2010;100% female literacy <40 by 2010

More ambitious rate of change.

Income Poverty

Country Current status MDG Target 2015 PRSP/national target Comparison of national target to MDG

Albania 17% extreme poverty ($1 per day) 1998

11% 9% (2015) More ambitious

Benin Baseline data lacking.

Halve % on<$1 per day. Halve % on<$1 per day. Same

Guyana 35% below national poverty line (1999)

26.1% (national, not $ per day) 31.4% (2005) More ambitious.

Rwanda 60% 30% <30% Same Tanzania 36% below basic

needs, 19% below food poverty line

19.3% basic, 10.8% food 17.8% basic, 9.3% food by 2010

More ambitious

Uganda 35% 28% 10% (2017) More ambitious Vietnam 40% reduction 2000–10 in

percent of population living below international poverty line, 75% reduction in those below international food poverty line.

26

Hunger and Malnutrition

Country Current status MDG Target 2015 PRSP/national target

Comparison of national target to HJ?/////MDG

Albania 8% malnutrition of <5s in 1999

4% (no 1990 baseline available) None

Benin 16% below minimum energy consumption, 23% <5s malnourished.

Halve % hungry and % malnourished

More ambitious reduction in underweight children.

Cambodia Halve, between 1990 and 2015, the proportion of people who suffer from hunger

More ambitious reduction in stunted <5s

Ghana 25% (1999) Halve, between 1990 and 2015, the proportion of people who suffer from hunger

More ambitious reduction in underweight <5s

Nicaragua

1998: 20% chronic malnutrition of under 5s.

Halve, between 1990 and 2015, the proportion of people who suffer from hunger

2004:16% Faster reduction

Tanzania Halve, between 1990 and 2015, the proportion of people who suffer from hunger

Reduce prevalence of stunting from 43.4% to 20%. Reduce the prevalence of wasting from 7.2% to 2%.

Vietnam 34% <5 malnutrition (2000)

Reduce<5malnutrition to 25% 2005, 20% 2010.

Faster reduction than MDGs (WDI base year)

IMR Infant mortality rate; CMR Child mortality rate; ARI Acute respiratory infection; <5MR Under-five mortality rate; NER Net enrollment rate. PCR Primary completion rate; UPE Universal primary education;.UN United Nations; WB World Bank; WDI World Development Indicators. Note: Where cells are left blank, we lacked the baseline data to calculate what the MDG target would imply for the 2015 target. a. Harrison, Klugman, and Swanson (HKS, 2004) classify the target as “more ambitious” based on the rate of reduction to 2004. Presumably, the 2015 target in the PRSP still does not achieve the MDG because of the need to make up ground lost when IMR increased after transition started. b. HKS say more ambitious based on target rate of improvement from 1999, but this follows slow progress in 1990s. c. See previous comment on IMR. d. HKS say more ambitious based on rate of change to 2005, but the 2010 target suggests that this rate of decline is not expected to be maintained thereafter. e. Sixty-nine percent from WDI data. f. HKS say less ambitious based on target of percentage completing in six years (more demanding than the MDG). Sources: National data and data base compiled by DECDG (Development Economics DataGroup), and used in Harrison, Klugman, and Swanson (2004).

27

APPENDIX C: OFFICIAL DEVELOPMENT AID

Table C1: What percentage of Official Development Aid is recorded in the balance of payments?

Official transfers and net disbursementsa as percentage of net ODA disbursementsb Country 1998 1999 2000 2001 2002

Albania 53 43 57 77 68 Benin 68 77 53 76 65

Burkina Faso 104 95 84 88 68 Cambodia n.a. n.a. 74 87 92 Ethiopia n.a. n.a. 57 54 72 Ghana n.a. n.a. n.a. 57 16 Nepal n.a. 107 90 75 70

Nicaragua 49 62 75 46 80 Rwanda n.a. n.a. n.a. 97 75

Tajikistan n.a. 33 35 74 114 Tanzania 36 79 87 40 0 Uganda 89 79 63 98 112 Vietnam 51 52 51 53 48 Average 64 70 66 71 73

Sources: a. International Monetary Fund b. Development Assistance Committee and World Development Indicators on donor-reported aid disbursements.

28

Table C2: What percentage of official flows supports government expenditure?

Net official flows in fiscal tables as percentage of net official flows

in the balance of payments Country 1998 1999 2000 2001 2002 2003 2004 Albania 91 105 87 64 82 43 64Benin 66 67 59 67 54 55 46

Burkina Faso 62 88 89 94 96 98 100Cambodia 0 0 63 53 61 57 54Ethiopia 0 0 82 93 88 108 99Ghana 0 0 0 155 86 83 80Nepal 0 51 63 64 61 57 76

Nicaragua 82 104 79 75 88 87 84Rwanda 0 0 0 80 71 72 79

Tajikistan 25 71 11 25 17 14 28Tanzania 92 48 61 87 73 77Uganda 78 88 89 75 96 98 95Vietnam 67 70 68 59 78 42 50 Average 70 77 68 76 73 68 72

Sources: See table C1.

Table C3: Budget Support as Percentage of Official Finance of Government Expenditure

Country

Budget support disbursements as percentage of gross official flows recorded in

“Government Revenues and Expenditures”’ 1998 1999 2000 2001 2002 2003 2004

Albania 30 13 22 4 16 15 8 Benin 23 30 2 34 5 12 n.a.

Burkina Faso 15 18 10 22 30 34 29

Cambodia n.a. n.a 16 7 16 13 10 Ghana n.a n.a n.a 37 43 52 56

Rwanda n.a n.a n.a 45 65 51 63 Tanzania 38 31 31 44 58 56 57 Uganda 36 34 32 42 52 56 50 Average 29 25 19 29 36 36 39

Note: Some data relate to financial years. FY2002/3 is reported as 2003. Data for 2004 are mostly program projections.

Sources: Latest PRGF reports, Government Revenues and Expenditures tables.

29

BIBLIOGRAPHY

Conway, Tim, Karin Christiansen, and Dan Lui. 2004. “What are the implications of the MDGs for the welfare of the poorest?” Overseas Development Institute (ODI) report for the “reaching the very poorest” policy team, U.K. Department for International Development (DFID), March 12, 2004. London: ODI. Unpublished Daima/ODI (Daima Associates and Overseas Development Institute). 2004. Joint Evaluation of General Budget Support to Tanzania, 1995-2004, Phase 2 Consultancy Report, September 2004. London,: ODI. Unpublished. Forum on Alignment Challenges in Programme-Based Approaches. 2003. “Enhancing Implementation Effectiveness of Programme-Based Approaches through Program and Stakeholder Alignment.” Document prepared for the meeting held on November 17–19, 2003, Berlin. Foster, Mick. 2004. “MDG-Oriented Sector and Poverty Reduction Strategies: Lessons from Experience in Health: Main Report.” Final draft, December 2004. Washington, D.C.: World Bank. ———. 2003. “The Case for Increased Aid.” Final Report to the Department for International Development. December 2004. Chelmsford: Mick Foster Economics Ltd. Foster, Mick, Patrick Chaussepied, and J. Shivakumar, 2004, Evaluation of the Public Expenditure and Financial Accountability Programme Consultancy Report, June 2004. Unpublished. Harrison, Makiko, Jenny Klugman, and Eric Swanson. 2004. “Are PRS’s Undercutting the MDGs? An Empirical Review.” April. Hutton, Guy. 2002. “User Fees and Other Determinants of Health Sector Utilization in Africa: A Review of Formal and Informal Health Sectors.” Swiss Development Cooperation and Swiss Tropical Institute. Final Draft August 27, 2002. IMF (International Monetary Fund). 2005. “The Macroeconomics of Managing Increased Aid Inflows: Experiences of Low-Income Countries and Policy Implications.” Policy Development and Review Department. Washington, D.C.: IMF, August 8, 2005. ———. 2004a. “External Grants and IMF Policies.” Washington, DC: IMF. ———. 2004b. “Follow Up on the Recommendations of the Independent Evaluation Office Report on Fiscal Adjustment in IMF-Supported Programs.” Policy Development and Review and Fiscal Affairs Departments. February 9, 2004. Imf.org/external/np/pdr/fa/2004/eng/020904.pdf ———. 2004c. “Follow Up on the Recommendations of the Independent Evaluation Office Report on Fiscal Adjustment in IMF-Supported Programs, Report of Executive Board Discussion.” Public Information Note, March 11, 2004. Washington, DC: IMF.

30

———. 2004d. “Are IMF-Prescribed Fiscal Targets Too Tight? IMF Survey 33(12). June 28. ———. 2003. “Aligning the PRGF and the PRSP Approach: Issues and Options.” Prepared by the Policy Development and Review Department. April 25, 2003. imf.org/external/np/prsp/2003/eng/042503.pdf IMF and IDA (International Development Association). 2003. “PRSPs—Progress in Implementation.” Prepared by the staffs of the IMF and the World Bank. September 12, 2003. Imf.org/external/np/prspgen/2003/091203.htm. IEO (Independent Evaluation Office). 2004. “Evaluation Report on PRSPs and the PRGF.” Washington, DC: IMF, July 7, 2004. ———. 2003. Fiscal Adjustment in IMF-Supported Programs. Independent Evaluation Office Report, September 8, 2003. Washington, DC: IMF. Jones, G., R.W. Steketee, R.E. Black, Z.A. Bhutta, and S.S. Morris. 2003. “How many child deaths can we prevent this year?” Lancet 362(9377): 65–71. Knippenberg, R., A. Soucat, and W. Vanlerberghe. 2004. “Marginal Budgeting for Bottlenecks: A Tool for Performance-Based Planning of Health and Nutrition Services for Achieving the MDGs, Concept Note.” Washington, DC: World Bank. Millennium Project. 2004. “Summary of a Global Plan to Achieve the MDGs.” Draft report for discussion at the Organization for Economic Cooperation and Development–Development Assistance Committee High-Level Consultation on the UN Millennium Project, New York, NY, July 8–9, 2004. Millennium Project. 2003a. “MDG Needs Assessments: Country Case Studies of Bangladesh, Cambodia, Ghana, Tanzania and Uganda. Working draft,” January 17, 2003. New York, NY: Millennium Project. ———.2003b. “MDG Country Case Studies: Methodology and Very Preliminary Results.” Draft, October 2003. New York, NY: Millennium Project. Mills, Anne. 2004. “Achieving the Health MDGs.” Presentation to DFID economist’s retreat, July 2004. Nkusu, Mwanza. 2004. “Aid and the Dutch Disease in Low-Income Countries, Informed Diagnoses for Prudent Prognoses.” IMF Africa Dept Working Paper. Washington, DC: IMF, March. OED (Operations Evaluation Department).2004a. “The Poverty Reduction Strategy Initiative: An Independent Evaluation of the World Bank’s Support through 2003.” Washington, DC: World Bank.

31

———. 2004b. “OED Review of the Poverty Reduction Strategy (PRS) Process, Mauritania Case Study.” Washington, DC: World Bank, OED, Corporate Evaluation and Methods Group. Processed.

Sachs, Jeffrey D. 2003. “Mobilization of Domestic and Donor Resources for Health: A Viewpoint, Increasing Investments in Health Outcomes for the Poor.” Second Consultation on Macroeconomics and Health, Geneva, October 29, 2003. Soucat, Agnes, Wim Van Leberghe, François Diop, Son Nam Nguyen, and Rudolf Knippenberg. 2002. “Marginal Budgeting for Bottlenecks: A New Costing and Resource Allocation Practice to Buy Health Results.” World Bank, Policy and Sector Resource team, Africa Region Human Development. Draft, November 2002. Washington, DC: World Bank. UNDP (United Nations Development Programme). n.d. “Frequently Asked Questions,”http://www.undp.org/mdg/faqs.shtml; accessed September 1, 2005. Wagstaff, Adam, and Mariam Claeson. 2004. The Millennium Development Goals for Health: Rising to the Challenges. Washington, DC: World Bank. World Bank. 2004a. Chapter 8 in World Development Report 2004: Making Services Work for Poor People. Washington, DC: World Bank. ———. 2004b. “Ethiopia, First Poverty Reduction Support Operation.” Program Document, January 23, 2004. Washington, DC: World Bank.

———. 2004c. “The PRSP Process, NEPAD and the Millennium Development Goals.” Informal paper, April 2004. Washington, DC: World Bank, Chief Economist’s Office, Africa Region.

———. 2004d. “Nepal, Project Appraisal Document, Health Sector Program,” August 4, 2004. Washington, DC: World Bank.

———. 2004e. “Targets and Indicators for MDGs and PRSPs: What Countries Have Chosen to Monitor.” Draft, September 4, 2004. Washington, DC: World Bank, Development Data Group. Unpublished. ———. 2003. “Supporting Sound Policies with Adequate and Appropriate Financing: Implementing the Monterrey Consensus at the Country Level.” Paper prepared for Development Committee meeting, September 22, 2003. Washington, DC: World Bank. ———. n.d. “Interim Guidelines for PRSCs.” OVC toolkit. Washington, DC: World Bank. http://info.worldbank.org/etools/docs/library/108875/toolkit/sector/cas.htm WHO (World Health Organization). 2004. “PRSPs: Their Significance for Health.” Second Synthesis Report. Geneva: WHO. ———. 2003. “Macroeconomics and Health, an Update: Increasing Investments in Health Outcomes for the Poor.” Second consultation on macroeconomics and health, Geneva, October 29, 2003.

H N P D I S C U S S I O N P A P E R

About this series...

This series is produced by the Health, Nutrition, and Population Family(HNP) of the World Bank’s Human Development Network. The papersin this series aim to provide a vehicle for publishing preliminary andunpolished results on HNP topics to encourage discussion and debate.The findings, interpretations, and conclusions expressed in this paperare entirely those of the author(s) and should not be attributed in anymanner to the World Bank, to its affiliated organizations or to membersof its Board of Executive Directors or the countries they represent.Citation and the use of material presented in this series should takeinto account this provisional character. For free copies of papers inthis series please contact the individual authors whose name appearson the paper.

Enquiries about the series and submissions should be made directly tothe Managing Editor Rama Lakshminarayanan (rlakshminarayana@worldbank.org) or HNP Advisory Service (healthpop@worldbank.org,tel 202 473-2256, fax 202 522-3234). For more information, see alsowww.worldbank.org/hnppublications.

THE WORLD BANK

1818 H Street, NWWashington, DC USA 20433Telephone: 202 473 1000Facsimile: 202 477 6391Internet: www.worldbank.orgE-mail: feedback@worldbank.org

MDG Oriented Sector and Poverty ReductionStrategies

Lessons from Experience in Health

Mick Foster

October 2005